We are studying how aberrantly silenced and DNA hypermethylated colorectal cancer (CRC) genes initiate early steps in CRC initiation and the molecular determinants of this gene silencing. The first aim focuses on SOX17, newly discovered to be DNA hypermethylated in ~90% of pre-invasive colon neoplasms. It encodes a protein essential for gut endoderm differentiation which inhibits, in the nucleus, Wnt pathway activation of 2-catenin mediated transcription. We are exploring whether direct SOX17 binding to gene promoters is involved, the genes targeted, and the role in CRC initiation. In our second aim, we explore mouse models for CRC evolution based on epigenetic loss of Hic1, a zinc finger, transcriptional repressor. Hic1 complexes with SIRT1 to directly suppress SIRT1 transcription. Hic1 deficient cells have decreased p53 function and resistance to DNA damage. Sirt1 may participate in silencing of DNA hypermethylated CRC genes. Hic1+/- knockout mice have increased colon Sirt1, abnormal colon crypt formation, and develop Hic1 negative, Sirt1 positive, colon polyps. Apc+/- (Min), Hic1+/- double het mice have markedly accelerated colon tumorigenesis. We are exploring how Sirt1 and other epigenetically silenced genes, may mediate this. The third aim utilizes siRNA and genetic approaches in primary, and/or immortalized, colonocytes, to explore how DNA hypermethylated mediated gene silencing, and the gene numbers involved, may help initiate CRC. Finally, we are asking whether increases in polycomb group (PcG) and associated proteins is a stress/survival response in CRC risk states, which may initiate abnormal epigenetic gene silencing. We explore a "molecular progression" model wherein these proteins recruit DNA methyltransferases (DNMT's) to gene promoters and the DNMT's, through transcriptional co- repression, may contribute to gene silencing prior to appearance of DNA methylation. We will time the appearance of these molecular events, including their localization to specific gene promoters, and will increase the protein complexes in DNMT deficient CRC cells to determine how genes are targeted for silencing. Relevance to Public Health: Our studies seek to identify how CRC risk states lead to abnormal silencing of genes. Reactivation of these genes holds promise for constructing prevention and treatment strategies for this disease. PUBLIC HEALTH RELEVANCE: A common event in tumorigenesis is abnormal, promoter, CpG island, DNA hypermethylation and associated epigenetic gene silencing. We investigate whether this change can foster cancer initiation and the molecular mechanisms though which it arises.